MGdaasLab/WHartTest

What it solves

WHartTest is an AI-driven intelligent automation testing platform designed to streamline the entire testing lifecycle. It solves the problem of manual, repetitive test case creation and maintenance by automating the generation, management, and execution of functional, API, and UI tests based on requirements documents and business context.

How it works

The platform uses a Monorepo architecture combining a Django backend and a Vue frontend. It leverages LangChain and LangGraph for multi-turn reasoning and orchestration, and the Model Context Protocol (MCP) for tool calling. It integrates RAG (Retrieval-Augmented Generation) via a knowledge base that vectorizes product and interface documents to provide context for the AI. The system includes a dedicated UI automation executor and an Agent skill library to perform actual test execution and analysis.

Who it’s for

Testing teams and QA engineers who want to automate the generation of test cases from requirements, reduce manual script writing for API and UI tests, and use AI to analyze failures and suggest fixes.

Highlights

• AI-Powered Case Generation: Automatically generates test cases from requirements, business descriptions, and knowledge base context.
• Comprehensive Test Coverage: Supports functional, API, and UI automation testing with AI-assisted writing and repair.
• RAG-Integrated Knowledge Base: Uses embedding models and rerankers to turn product documentation into actionable AI context.
• Agentic Workflow: Employs LangGraph-based agents that can call external tools via MCP and specialized skills (e.g., Playwright) for execution.
• Intelligent Failure Analysis: Automatically analyzes execution logs, screenshots, and traces to locate problems and generate repair suggestions.